Title
!pip install -Uqq fastbook
import fastbook
fastbook.setup_book()
from fastbook import *
from fastai.vision.widgets import *
Azure bing search resource setup
key = os.environ.get('AZURE_SEARCH_KEY', 'e1684f9aff0948c490fe0a1fec990e8d')
types = 'hotdog','random'
path = Path('hotdogs')
Folder creation and image search plus download
if not path.exists():
path.mkdir()
for o in types:
dest = (path/o)
dest.mkdir(exist_ok=True)
results = search_images_bing(key, f'{o}')
download_images(dest, urls=results.attrgot('contentUrl'))
Test downloaded files and remove failures
fns = get_image_files(path)
fns
failed = verify_images(fns)
failed
failed.map(Path.unlink);
hotdogs = DataBlock(
blocks=(ImageBlock, CategoryBlock),
get_items=get_image_files,
splitter=RandomSplitter(valid_pct=0.2, seed=42),
get_y=parent_label,
item_tfms=Resize(128))
dls = hotdogs.dataloaders(path)
dls.valid.show_batch(max_n=4, nrows=1)
Train the Model
hotdogs = hotdogs.new(
item_tfms=RandomResizedCrop(224, min_scale=0.5),
batch_tfms=aug_transforms())
dls = hotdogs.dataloaders(path)
learn = cnn_learner(dls, resnet18, metrics=error_rate)
learn.fine_tune(4)
interp = ClassificationInterpretation.from_learner(learn)
interp.plot_confusion_matrix()
interp.plot_top_losses(5, nrows=1)
Data Cleaning
cleaner = ImageClassifierCleaner(learn)
cleaner
for idx in cleaner.delete(): cleaner.fns[idx].unlink()
for idx,cat in cleaner.change(): shutil.move(str(cleaner.fns[idx]), path/cat)
Exporting
learn.export()
path = Path()
path.ls(file_exts='.pkl')
learn_inf = load_learner('export.pkl')
img = PILImage.create('hotdog.jpg')
out_pl = widgets.Output()
out_pl.clear_output()
with out_pl: display(img.to_thumb(128,128))
out_pl
pred,pred_idx,probs = learn_inf.predict(img)
lbl_pred = widgets.Label()
lbl_pred.value = f'Prediction: {pred}; Probability: {probs[pred_idx]:.04f}'
lbl_pred
btn_run = widgets.Button(description='Classify')
btn_run
def on_click_classify(change):
img = PILImage.create(btn_upload.data[-1])
out_pl.clear_output()
with out_pl: display(img.to_thumb(128,128))
pred,pred_idx,probs = learn_inf.predict(img)
lbl_pred.value = f'Prediction: {pred}; Probability: {probs[pred_idx]:.04f}'
btn_run.on_click(on_click_classify)
#Putting back btn_upload to a widget for next cell
btn_upload = widgets.FileUpload()
VBox([widgets.Label('Select your hotdog!'),
btn_upload, btn_run, out_pl, lbl_pred])